About this skill

Why this exists

Useful signals in Douyin comments are often scattered across first-level comments, replies, and informal phrasing. When analyzing competitive feedback, support FAQs, product pain points, or public sentiment, manual pagination can miss low-frequency objections. This skill turns comments under a known video or share link into readable, structured results, then groups them into audience feedback themes.

How it works

The skill is built around SocialDataX data calls and uses SOCIALDATAX_API_KEY for requests. When the agent can run shell commands, prefer the direct CLI: anchor a known video or post with --aweme-id, pass a content page URL, short link, or share text with --url, and combine --aweme-id with --comment-id for replies. Pagination relies on --page-token; return the exact next_page_token from the previous page and do not reuse tokens across content items or first-level comments.

If MCP tools are already available, use douyin_get_video_comments_by_aweme_id, douyin_get_video_comments_by_url, or douyin_get_video_comment_replies_by_comment_id. Before inferring sentiment or demand, group comments by observed themes, note whether the result is one page or multiple pages, and treat empty comments as a valid successful result.

Boundaries

The skill is read-only: it does not read local browser data, save API keys, or perform login, posting, liking, commenting, or account changes. Do not pass playback URLs such as video.play_url as content page URLs, and do not pass both --aweme-id and --url to the same comments command. It is suited for comment insights, feedback extraction, and discussion summaries, not for private account scraping or replacing compliance review.

Screenshots

Use Cases

During competitive analysis, read first-level comments on a Douyin video and summarize complaints about price and delivery.
While preparing support materials, extract high-frequency FAQ candidates from hot-video comments and replies.
During product review, pull comments for target videos from share links and compare pain points and objections.
During sentiment monitoring, group first-level comments for a viral video and flag emotional tone and disputed points.

Best For

Product managers doing competitive feedback: need to summarize pain points, objections, and demand signals from target video comments.
Customer support operators: need to extract high-frequency questions from hot-video comments and replies into FAQ or response drafts.
Content analysts: need to group first-level comments by theme and assess sentiment or discussion focus.
Agent developers: want to integrate read-only Douyin comment data via CLI or MCP for downstream analysis.